Would changing the algorithm be something feasible and that would allow more security? (To the end of making it easier for people to mine this coin and support the network/coin.) I don't personally have an ASIC... I can only purchase bitcoin and use Nicehash or something if I want to mine this algo with any effectiveness. Whereas I've found algorithms such as cryptonight, cryptolight, qubit, nist5, lyra2rev2, Zcoin's algo, and a couple others to be pretty easy to GPU mine). Some coins with those algorithms have quite sizable network hash rates.
Obviously noone's mining CHAO lately (the block explorer shows more than three days since a block was mined), so if the mechanism described in the Cryptopia quote is what occurred it is likely to happen again if there's not a sustained hashrate applied to this coin. 23-Skidoo seems kind of a victim of there being too many other coins out there and not a lot of attention paid to maintaining it (and the circumstance is somewhat bizarre because it's far scarcer than Unobtanium but the latter seems to be thriving).
This situation also points out the dangers of having too much coin parked at an exchange w/o at least a portion in a wallet under one's personal control/agency (and, in particular, a paper wallet for some of it). Apparently an exchange can be an obvious centralized target for thieving cretins.
There is no problem with algorithm. The problem is reward method. Nobody will mine a coin that has 0 block reward and 0 transactions in that block (means 0 fees).
Solutions are:
1) increase block rewards, I think can remain different block reward amount but increase it, for example: 250 sat / 500 sat / 750 sat / 1000 sat / 1500 sat (instead of 0/1/10/100/1000). After pay relay fee the mined amount will be 50 / 300 / 550 / 800 / 1300
2) decrease relay fee. Last prices of the coin are very high, so with relay fee 1000 sat the coin has more tx fee than bitcoin, nobody thinks that is not ok?